In this post I shall:
- Explain why deposit insurance, as currently employed and lauded almost everywhere, is a deeply flawed and damaging idea, simultaneously destabilising capitalism, necessitating invasive anti-innovating regulation, and encouraging regulatory laziness.
- Set out a new alternative form of deposit insurance, which would imply a materially different structure of the banking system from what has gone before.
Here's the summary. Feel free to comment on the basis of the summary - the main piece is long.
Summary
Conventional deposit insurance is bad because:
- Without invasive regulation it encourages a damaging process of escalation in riskiness of the financial system, as higher-risk banks suck deposits out of low-risk banks until all banks are high-risk. With invasive regulation, innovation and risk diversification are reduced, along with growth in the economy.
- In the traditional British regulatory model, with an all-powerful Bank of England, deposit insurance encourages laziness on the part of the regulator, and leads to more bank failures.
As an alternative to conventional deposit insurance I propose
- Insurance of cheque accounts into which salaries are paid electronically, up to a model limit such as £10,000.
- Insurance of a new type of bank account, which would be fully backed by the bank buying government gilts, and the deposits of which could not be used in fractional reserve banking (standard banking). This would be as close as it gets to a risk-free product.
Main Piece
Before I move to my specific critiques of deposit insurance, I want to begin by offering you a much vaguer thought, to convince you that, even if you don't completely follow or completely accept my later specific critiques, nonetheless you should feel that something about deposit insurance "smells bad".
The standard way that banks work is called "fractional reserve banking". What this means is that banks take deposits and lend out large multiples of those deposits held at any one time, keeping only a small reserve of funds available to pay out to people withdrawing deposits. The consequence is that banks "owe" their depositors ten times, fifty times, perhaps a hundred or more times as much money as the bank has available for instant withdrawal. If all depositors attempt to take out their money at the same time, then a bank that does fractional reserve banking will fail, even if it is a profitable and otherwise solvent business. Thus, fractional reserve banking is an intrinsically and ineliminably risky activity.
People depositing their funds in banks provide capital that the banks then use to engage in lending at risk. They are providers of capital, like shareholders or bondholders (and indeed deposit funds can be and have been replaced by alternative sources of capital). Banks pay depositors deposit interest rates higher than the amount depositors would receive if, instead of lending their money to the bank (which is what a deposit is - a form of lending), they put their money into government bonds (the closest thing to risk-free lending that there is).
So, people lend money to intrinsically and ineliminably risky institutions that then use that money to engage in risky activities. Deposit insurance appears, perhaps, to remove some of that risk by saying that if everyone attempts to take out their money at the same time and the bank fails, then they will still get their money back. But if depositing in intrinsically risky institutions is then insured by the state and treated by as if it were risk-free, does that not seem as if it must be problematic, as if, somehow (even if you can't see how straight away) there is the danger of introducing a form of instability right at the heart of the capitalist system - the banks? Doesn't something not "smell right" here?
You should feel like that. Deposit insurance creates a significant problem, and undermines a key solution to other problems, as follows. (You will have one large pragmatic objection as I go along - be patient and I hope I shall satisfy you at the end.)
First, deposit insurance necessitates invasive and restrictive regulation of financial innovation. Why? Well, suppose that we begin with all the banks having low-risk business models and we have only light regulation. Then someone produces an innovation - a new bank that engages in slightly riskier activities than other banks and consequently has slightly higher expected returns. Because the returns to its activities are higher, this bank can afford to pay slightly higher deposit interest rates. The depositor doesn't care whether her money is in the low-risk or higher-risk bank, because it is insured by the state, and so takes advantage of the higher interest rate by switching deposits to the higher-risk bank. Consequently the low-risk banks become insolvent and disappear (no-one deposits there any more), leaving only the higher-risk banks. Then there is another innovation - a yet more risky bank, paying yet higher deposit rates. And so it goes on with the banks becoming riskier and riskier, until eventually there is a system-wide shock and almost all the banks become insolvent at once.
This process means that if we employ deposit insurance, eliminating the incentive for depositors to take an interest in the riskiness of banks, we must pair it with regulation that restricts the activities of banks. If we don't want that restrictive regulation - limiting innovation and reducing the ability of financial markets to innovate to meet our needs and diversify away our risks - if we don't want that regulation we must oppose the deposit insurance that makes it inevitable.
Second, deposit insurance undermines the best form of prudential regulation. There are two key schools of prudential regulation through banks. I call them the "monarchical" system and the "written constitution" system. The old Bank of England regulatory framework applied the monarchical system. The Bank was that monarch, empowered to do whatever it took to deliver its mission. If a bank were solvent but faced a temporary liquidity problem arising from a systemic issue, then the Bank would provide liquidity. If a bank were insolvent or were going to become insolvent, then the Governor of the Bank would take the controllers of another bank down to his club and explain to them that they were going to take over the insolvent bank - and like it. Then he would explain to the insolvent bank that it was going to be taken over - and like it. And then the next day it would be announced publicly that the insolvent institution were being taken over by the other institution - and everyone would like it. Earlier than that, the monarch Bank would inspect the banks under its care, and if something didn't seem quite ship-shape and Bristol fashion, or not quite cricket, then there would be a quiet word in the ear and the bank would sort itself out. The monarch Bank had the tools and the authority to do what was necessary.
An alternative regulatory model, the "written constitution" model, involves the setting out of a set of rules specifying how much of this or that sort of thing banks are allowed/required to do on this or that occasion. Now, of course, one cannot write down a rule that tells you the best thing to do in all circumstances - there is the problem of unforeseen contingencies. Consequently, under any rules-based system it is possible that a bank that actually has a sound business model is inadvertently shut down as a result of the rules. Similarly, there is always the possibility that something comes along that no-one thought of, such that rules that worked perfectly well on most occasions fail on these. Either way, sound institutions might occasionally go bust as a by-product of or through inadequacy of the rules.
Under the written constitution model it is appropriate (indeed necessary) to employ deposit insurance of some sort - the inflexibility of the system does not otherwise properly protect those that deposited in perfectly sound institutions, and willingness to deposit at all will be reduced. On the other hand, under the monarchical model the question of compensating depositors in solvent institutions doesn't arise - the flexibility of the monarchical system allows the Bank to prevent solvent institutions from failing.
We see some of the weakness of the rules-based system in the current crisis. Such a scenario is simply not envisaged in the Basel II framework or the risk-assessment models arising from it. Employing hindsight, some suggest we should now include liquidity provisions in Basel II. Fine, but that won't allow the rules to deal with next-time's unforeseen contingency.
In such circumstances there would have been considerable advantages in having a monarchical Bank of England - such an arrangement allowed the UK to navigate many an international financial crisis over decades with relatively little financial turmoil here and this stability was an important reason for the strong presence of financial services in the UK.
What will happen if we introduce deposit insurance into a monarchical model? Answer: the regulator will adjust for the presence of deposit insurance, in a process akin to cyclists adjusting to the presence of a helmet by taking more risks or cigarette smokers compensating for lower tar levels by smoking more and inhaling more deeply. Deposit insurance allows a monarchical regulator to be laxer, allowing a few extra firms to collapse rather than be provided with liquidity or taken over, with activities monitored that little bit less - because the public is protected in the end. Deposit insurance leads to more bank failures under a monarchical system, not fewer.
So, I have argued that deposit insurance necessitates invasive regulation and results in more bank failures. And yet, and yet...I am sure that many readers will want to say: But people depositing their funds in a bank don't want to work out how risky its activities are anyway! So whether there is deposit insurance or not there will be the risk-escalation process you describe. And do you think it is really politically feasibly to allow old grannies to lose all their life savings when their bank down the road fails just because the granny didn't adequately monitor its activities?! And what about people who just have their pay put into their accounts electronically? Of what interest is it to them what other activities the bank engages in? Are they supposed to get their wages in notes in a brown envelope?!
Hence, here are two forms of something akin to "deposit insurance" that, as Montgomery Burns might have said, I don't hate.
- Insurance of one chequing account into which salary payments are made. I suggest that a ceiling of £10,000 should be quite adequate, and would impose the regulatory requirement that zero interest be paid on insured accounts.
- A requirement on all banks licensed to engage in fractional reserve banking that they provide a special sort of deposit account that I shall term a "gilt aggregator account" (GAA). Such an account is effectively the purchase of a share in a government gilt fund, and each pound deposited in the account must be fully backed by one additional pound of government gilt purchase. Banks would pay the government gilt rate they themselves receive minus an administration fee (built into the "deposit" rate paid). They would make their money on this business through the economies of scale in transactions costs versus someone buying their own personal gilts. Funds deposited into GAAs would be excluded from regulatory requirements in respect of fractional reserve banking - they could not be offered as capital reserves - and the gilt fund backing would be legally ring-fenced from the other assets of the bank in the event of bank failure. Since these funds are, as near as can be, risk-free, the government insurance really addresses situations such as fraud or gross incompetence when the gilt backing does not work.
Under such a system, people would be able to engage in electronic payments and risk-free storage of their money. Of course, gilt aggregator accounts would pay very low interest, so banks would be able to offer attractive time deposit accounts paying much higher rates of interest, but would have to warn depositors that if they switched their monies out of the GAA then they would no longer be insured by the state. I believe that in this event, with such clear warnings, people would understand that monies on bank deposits outside GAAs are lending-at-risk, and that depositors are capital providers, investors like bondholders or shareholders. Of course, it is likely that, with state insurance totally eliminated, banks would seek to obtain private insurance of their deposits, and would be at liberty to quote their private insurance to potential depositors.
This scheme could not be introduced immediately. Under current circumstances it would imply a dramatic contraction in banks since monies moved into GAA deposits would be excluded from regulatory prudential capital. But over the longer term, I suggest that this is the way to go, rather than continuing down the high-regulation path of conventional deposit insurance.
